Selective Accumulation of the Complement Membrane Attack Complex in Aging Choriocapillaris

The membrane attack complex of complement (MAC) is a multimeric assembly of proteins consisting of C5b, C6, C7, C8 and multiple copies of C9, which forms the membrane spanning pore (Morgan, 2015). Although the function of MAC is to lyse pathogens, aberrant activation can lead to bystander injury of resident cells (Rus and Niculescu, 2001). The MAC is abundant along the walls of kidney glomerular capillaries in an experimental model of glomerulonephritis, and absent in healthy control samples (Koffler et al., 1983). In contrast, even healthy human maculae labeled with anti-MAC antibodies exhibit some degree of labeling in domains surrounding the choriocapillaris, but little or no labeling in the normal RPE or neural retina (Hageman et al., 2005; Seth et al., 2008). During normal aging, the levels of MAC increase in abundance. This is especially notable in the choroids of individuals with high-risk complement factor H (CFH) genotypes (Mullins et al., 2011a) and patients with age-related macular degeneration (AMD) (Mullins et al., 2014).

The molecular basis for MAC activation in the choroid is not fully understood, although it may be a result of the unusual composition of the outer retina (such as long chain fatty acids, phospholipids and abundant bisretinoids (Curcio et al., 2011; Zhou et al., 2009)) which creates a pro-inflammatory choroidal microenvironment. While robust anti-MAC labeling is observed in aging choriocapillaris, whether this deposition is choroid-specific has not been systematically examined. In the current study we sought to examine the abundance of the MAC in the microvasculature of a variety of human tissues to aid in our understanding of the role of its deposition in aging and AMD.

Human donor eyes were acquired from the Iowa Lions Eye Bank (Iowa City, IA) following informed consent of the family and in accordance with the Declaration of Helsinki. Posterior poles were fixed in 4% paraformaldehyde in phosphate buffered saline, and full thickness (retina to sclera), 4mm-diameter punches were collected from six donor eyes (ages 41 to 83) without any known retinal disease adjacent to the macula. Ocular tissue was processed at the University of Iowa Central Microscopy Research Facility using an RMC Paraffin Tissue Processor 1530 to dehydrate and infiltrate with paraffin before embedding in a paraffin block. Sections were cut at 5 μm using a microtome and collected on charged glass slides. Tissue array slides with 38 healthy tissue samples from 19 different anatomic sites were obtained from a commercial vendor (Pantomics; MNO381; Richmond, CA). These tissue arrays contain sections from the following tissues: breast, cerebellum, cerebral cortex, colon, esophagus, heart, kidney, liver, lung, ovary, pancreas, prostate, skin, small intestine, stomach, testis, thyroid, tonsil, and uterus. For all procedures, from deparaffinization through final coverslipping, ocular sections and tissue array sections were handled identically.

Sections were deparaffinized and rehydrated prior to staining. Immunohistochemistry was performed with antibodies directed against the MAC (M0777; Dako, Carpinteria, CA) or ICAM-1 (BBA17; R&D, Minneapolis, MN; included as a positive control to verify preservation of antigenicity), or were labeled with the fucose-binding lectin Ulex europaeus agglutinin-I (UEA-I; Vector Laboratories, Burlingame, CA).

Immunohistochemical labeling and colorimetric detection (using avidin-biotin horseradish peroxidase and VIP substrate, both from Vector Laboratories) were performed as described previously (Skeie et al., 2010).

Of the 19 tissue types evaluated, the choriocapillaris was the only tissue with abundant MAC immunolabeling (Figure 1A), which was absent from all other tissues (Figure 1E, I, M, Q). ICAM-1 was present in the vascular endothelial cells of all tissues except stomach, which had milder staining overall (Figure 1G), and was highly expressed in choroid (Figure C) and kidney glomerulus (Figure O). Small intestine exhibited strong intracellular epithelial cell staining (Figure 1K), and the pancreas tissue showed high intracellular pyramidal cell staining, with minimal vascular labeling (Figure 1S). UEA-I lectin labeling was detected in all tissues on the vasculature (Figure 1B, F, J, N, R), with some tissues showing labeling in specific cell classes (for example, gastric mucosa (Al-Marzoqee et al., 2012)). A summary of the data can be found in Table 1.

One limitation of this experiment is that choroid was not present on the commercial tissue arrays, and the processing of these tissues was outside of our control and therefore could result in different levels of antigenicity between ocular and extraocular tissues. We sought to address this potential problem by embedding our ocular tissue in paraffin, performing all labeling steps for ocular and tissue array samples together, and including positive controls for preservation of protein and carbohydrate epitopes (anti-ICAM-1 and UEA-I), both of which showed strong labeling in both ocular and tissue array sections.

In order to further assure that the differences in MAC labeling between choroid and the other tissues studied were not due to fixation and processing, we performed a second set of comparisons in which we obtained unfixed sections of human tissues from the University of Iowa Tissue Procurement Core Facility and compared them with similarly aged unfixed donor eye sections. Sections were collected, labeled with anti-MAC antibody and biotinylated UEA-I lectin, and evaluated by fluorescence microscopy as described previously (Mullins et al., 2014). As observed with the first comparison, the choriocapillaris had the most abundant MAC labeling between all the tissues (Figure 2A, A′). While low levels of the MAC were present within kidney (Figure 2C, C′) and liver (Figure 2D, D′), the microvasculature in all 7 extraocular tissue sections were negative for the MAC (Figure 2B, B′, E, E′, F, F′, G, G′, H, H′).

In summary, the selective accumulation of MAC in the choriocapillaris is a plausible explanation for the fact that individuals with high risk CFH genotypes—which lead to increased systemic complement activation (Hecker et al., 2010; Reynolds et al., 2009)—develop macular degeneration rather than an array of extraocular diseases. For reasons that are not yet clear, the choroid appears to be a “hot spot” for MAC deposition. The impact of ocular specific factors in AMD pathogenesis, as opposed to systemic factors, is indicated by elegant studies of genotyped liver transplant patients, in which recipient CFH genotype, but not donor CFH genotype, contributes to AMD risk (Khandhadia et al., 2013).

The increased accumulation of MAC in AMD occurs concomitantly with the microvascular injury that occurs in individuals with AMD. Morphometric and molecular studies by us (Mullins et al., 2011b; Whitmore et al., 2013) and others (Biesemeier et al., 2014; Ramrattan et al., 1994) suggest that choriocapillaris loss is one of the earliest anatomical events in AMD, and that the MAC may represent a smoking gun for early AMD (Whitmore et al., 2014). Efforts are underway to employ anti-complement therapies for AMD (Birke et al., 2013; Rohrer et al., 2009). These findings suggest that identifying the molecular basis for choroid specific MAC activation is an important question, and they further support the notion that local inhibition of the terminal complement pathway will be necessary in delaying or arresting damage to the choriocapillaris and, ultimately, RPE and photoreceptor cells, in AMD.